Intermediate · 30 minYou are hereRead
Logistic regression
A weighted sum becomes a probability through the S-shaped sigmoid; the boundary at one half is a straight line, and the size of the weights sets how sharp it is.
Classical machine learning
Classification
A classifier gives each example a class, or a probability for each class. This topic turns a weighted sum into a probability with the sigmoid, scores it with cross-entropy, extends it to many classes with the softmax, and classifies by Bayes' theorem with naive Bayes.
Start with Logistic regression0 of 4 read
Lessons on this topic
Related topics